Research on the typical parallel processing algorithm of massive remote sensing data

With the rapid development of the earth observation technology, remote sensing image data is exponential growth, which has a huge challenge to the existing remote sensing processing system. Because of the limitation of single machine processing capability, the processing speed of single machine algorithm has been greatly improved. So the high performance parallel clustering technology is used. In this paper, the characteristics of remote sensing image geometric features, complex features and large data processing requirements of the algorithm are discussed. The parallel processing technology is studied. The parallel architecture and parallel computing models are discussed. The relationship between the number of threads and the number of cores is discussed. The experimental results are discussed. The running time of the system is influenced by the number of threads and the hardware.